Abstract—Cardiovascular diseases remain the leading cause of mortality worldwide, emphasizing the need for accurate early prediction. With the increased accessibility to clinical datasets, Machine Learning (ML)-classification algorithms and Deep Learning (DL) approaches (Convolutional Neural Networks) have become integral to early prediction of cardiovascular diseases. This work proposes an effectual hyperparameter optimized heart disease prediction system using a Multilayered Perceptron (MLP) and compares its performance with other Machine Learning (ML) algorithms. The Cleveland, Statlog and Hungarian heart disease datasets sourced from the UC Irvine Machine Learning Repository (UCI ML) were used. Initially, data pre-processing includes correlation analysis between the relevant body values and disease-followed by applying Synthetic Minority Over-sampling Technique (SMOTE) on the training folds only to address the class imbalances without affecting the data integrity. A 10-Fold Cross-Validation approach was used to ensure the robust model evaluation. Hyperparameter optimization using GridsearchCV further enhanced the generalization. Optimized MLP attained the highest accuracy levels of 96.89 followed by the Random Forest (96.07%), Decision Tree (94.99%), K-Nearest Neighbor (KNN) (94.10%), Support Vector Machines (SVM) (93.48%) and Logistic Regression (LR) (93.55%). The Findings obtained illustrate the efficacy of MLP architecture in the prediction of cardio vascular diseases for clinical applications.
 
Keywords—cardiovascular diseases, machine learning, hyperparameter optimization, random search, grid search, artificial neural networks, k-fold cross validation, Synthetic Minority Over-sampling Technique (SMOTE)


Copyright © 2026 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).